Make Your Own Solar Storms!
Overview
Solar storms, including coronal mass ejections (CMEs), are large explosive events on the Sun that are capable of ejecting billions of tons of magnetized, million-degree plasma into space. When CMEs reach Earth, they can disrupt radio transmissions, damage satellites, and severely impact power transmission grids, leading to extended, large-scale power outages. Predicting such eruptions and their properties is therefore of great importance and a major goal of NASA’s Heliophysics Research Program.
Project Details
Due to their inherent complexity, the mechanisms that control CME properties are poorly understood and challenging to model empirically. Therefore, data-constrained magnetohydrodynamic (MHD) simulations of CMEs are a promising path forward for improving our understanding and advancing space weather prediction capabilities. To facilitate simulations for use by the space weather and heliophysics scientific communities, we have developed an interactive modeling framework called CORona-HELiosphere (CORHEL)-Automated Multiple CME Generator (AMCG) that allows non-expert users to routinely model multiple CMEs in a realistic coronal and solar-wind environment and propagate them to Earth. The CORHEL-AMCG framework is designed to run on high-performance computing platforms as well as in-house multi-GPU servers. This is a step towards future operational space-weather forecasting and rapid turnaround for research-focused CME simulations.
The CORHEL-AMCG framework consists of two main components. The first, AMCG, is a web-based interface that is used to design one or more pre-eruptive magnetic configurations, full MHD simulations that generate the background state of the solar atmosphere out to Earth, and high-fidelity CME simulations from the Sun to Earth. All simulations are performed with the second framework component: a heavily updated version of the CORHEL modeling suite, which consists of a collection of tools and simulation codes—including the Magnetohydrodynamic Algorithm outside a Sphere (MAS) code—linked together for ease of use. It also includes auto-generated, web-based visualization reports for each stage of the CME design and simulation.
Results and Impact
We have developed the CORHEL-AMCG framework and have used it to successfully model real solar storm events involving multiple eruptions. These event simulations were then used to model and study solar energetic particle events, which can pose a danger to astronauts in interplanetary space.
Why HPC Matters
The direct simulations of solar storms in CORHEL-AMCG require running moderate-to-large simulations with grids of up to 50 million cells and integrating time-dependent equations for millions of cycles. We therefore require the use of high-performance parallel computations. A typical sequence of a CORHEL-AMCG simulation, using the Intel Skylake nodes on NASA’s Electra supercomputer, is able to model twisted field design eruptions and the solar atmosphere background on eight nodes in approximately two and four hours, respectively. The solar storm eruption and propagation simulations require around 20 hours on 32 Electra nodes to complete. We found that the first two simulations are small enough to run on a single V100 GPU node on NASA’s Pleiades supercomputer, with the twisted field design eruption completing in under an hour and the solar atmosphere background taking approximately six hours.
What's Next
The CORHEL-AMCG framework has been delivered to NASA’s Community Coordinated Modeling Center, where it will be integrated into a public run-on-demand system, allowing researchers to routinely generate solar storm simulations from the Sun to Earth. This routine modeling of real events will allow us to learn much more about them, leading to improvements in space weather predictive capabilities.
More Information
- Magnetohydrodynamic Algorithm outside a Sphere (MAS) code
- Community Coordinated Modeling Center
- Sun-to-Earth MHD Simulation of the 14 July 2000 “Bastille Day” Eruption
Researchers
- Ronald Caplan, Predictive Science Inc. Please enable Javascript to see email address.
- Tibor Török, Predictive Science Inc. Please enable Javascript to see email address.
- Andres Reyes, Predictive Science Inc Please enable Javascript to see email address.
- Cooper Downs, Predictive Science Inc. Please enable Javascript to see email address.
